Text Transmit

Inspiration

With the prevalence of online learning, many applications have come to rise for making submission and assignments easy. But not many solutions exist for students to help share notes with each other, which is why we built Text Transmit.

What it does

With our current features, Text Transmit scans a document, and detects text with a neural network and computer vision. It then converts the words to a computer readable piece of text, which can be freely edited in the case of inaccuracies. Once you're done, you can save the note to the cloud and access it later.

How we built it

We settled on an iOS application, with our cloud saves powered by Firebase because that's what we had the most experience in. We used Apple's Vision API for text detection so all text is processed locally, with no risk of data breaches. Then, it is stored securely in Firebase.

Challenges we ran into

Creating the front-end UI was the hardest part of the process. Data transfer between views of our app was a particularly challenging task. Besides this, it was a bit challenging to set up the computer vision, but in the end that worked out.

Accomplishments that we're proud of

We have an (almost) fully functioning product after only 24 hours. Our project, while not as feature rich as we hoped it to be at the start of the hackathon, is pretty useful.

What we learned

Extensive planning, good collaboration, and research are all important when creating a project, especially within a time limit. There are some cases where if we bettered our communication, we could have gotten certain aspects of development done a lot faster, and had time to add additional features.

What's next for Text Transmit

We plan to fully implement viewing previous notes, sharing them with a specific set of users, and adding full document scanning, but don't have a timeframe for this.

Note: our demo on github requires some changes to the files, as the bundle id needs to be edited and cocoapods must be installed for testing. Our Firebase database will only connect to a specific bundle id, so the app's save features will not be functional.